Application of the neural network approach to predict the efficiency of the use of ecological neutrality transition technologies by power plants

Abstract

The issues of improving the environmental friendliness of power plants are relevant in Russia and in the world. The objectives of reducing greenhouse gas emissions and reducing heat losses in cooling towers are important. Reducing greenhouse gas emissions is relevant for both the urban economy and public health. Reducing thermal energy emissions from power plants within the city is also relevant, because. improves the quality of life of residents and guests of the city, reduces the load on air conditioning in summer, allows you to direct the released electrical energy to the development of the urban economy. There are many technologies that can be used to effectively utilize the thermal energy of cooling towers. The article presents those of them that have shown their technical and economic feasibility. The task of predicting the effects for a particular plant is complicated by the need to temporarily shut down part of the generating capacity, which is unacceptable for the city’s power supply system. Therefore, predictive models based on artificial neural networks are becoming the main tool. The assumptions used in solving problems of this class are shown, an example of a set of input factors is given, and the resulting errors are shown. It is shown that the combination of technologies for the transition to environmental neutrality of power plants can become the center of attraction for many urban projects that can increase economic efficiency, increase the number of jobs, develop import-substituting industries, and increase the efficiency of land use around generation facilities that were previously operated insignificantly.